Between 1978 and 1981, the NIMH Collaborative Depression Study conducted thorough baseline evaluations of 955 probands with major affective disorder. These subjects have now been followed at semiannual and annual intervals to ten years with a retention rate of 70.1% of those still living. The study also evaluated a large sample (N=2,616) of nonclinical subjects (relatives, controls and spouses) at baseline and again six years later. A high-intensity substudy has shown that the occurrence and timing of major depression disorder episodes among the members of this sample can be validly compared to corresponding measures in the proband sample. The size, scope, intensity, and duration of these follow-up efforts have thus generated a unique data set with very broad potential. Previous CDS publications have been based on shorter follow-up periods and have emphasized the course of index episodes, the time to subsequent clinical events, and the long-range prognostic qualities of baseline measures. The investigators now propose to fully explore the ten-year follow-up data with a shift in emphasis to the timing and phenomenology of prospectively observed episodes, to overall trends across repeated observations, to direct comparisons between probands and nonclinical subjects, and to the application of new statistical means toward these ends. The first three of these general aims will circumvent important selection biases which are inescapable for essentially all other follow-up studies. The last of these aims will be abetted through an affiliation with the biostatistics division in the Department of Preventive Medicine. The investigators intend an emphasis on nosological questions, a longstanding focus for researchers at the Iowa center.